Instability after a few hours.

[slicedtoad]

My rig is now water cooled and my temps are great. I'm almost stable at 4.7GHz but I get bsods after 8 or so hours of prime. Adjusting the voltage isn't helping. I've gone from ~1.410 to 1.450 and it always crashes around the 6-8 hour mark.

Temps don't go above 70C on the highest core but I've noticed my northbridge get's really hot. I don't have any accurate way of testing it's temperature but it's almost too hot to touch when running prime. My ram and voltage controllers (thingees by the cpu) get hot too but not nearly as hot.

What's the chance that water cooling my nb would solve my problems?

Btw, i'm not concerned about over volting my cpu. If it breaks i'll buy a new one (or grab an ivy bridge one).

 

[saint19]

Your can try with lower frequency.

 

[slicedtoad]

lower cpu frequency? that would defeat the purpose... I just want to know if cooling the nb can actually be useful.

 

[aqe040466]

My rig is now water cooled and my temps are great. I'm almost stable at 4.7GHz but I get bsods after 8 or so hours of prime. Adjusting the voltage isn't helping. I've gone from ~1.410 to 1.450 and it always crashes around the 6-8 hour mark.

Temps don't go above 70C on the highest core but I've noticed my northbridge get's really hot. I don't have any accurate way of testing it's temperature but it's almost too hot to touch when running prime. My ram and voltage controllers (thingees by the cpu) get hot too but not nearly as hot.

What's the chance that water cooling my nb would solve my problems?

Btw, i'm not concerned about over volting my cpu. If it breaks i'll buy a new one (or grab an ivy bridge one).

What temp monitor are you using? Download HW monitor and CPU-z in CPUID.COM

 

[slicedtoad]

im using core temp. hw monitor gives the same temps (and nothing gives nb temps). The voltage reading is from cpuz.
cpu heat isn't the problem though, it never goes above 70C. Before wc, I could run it at 85C stable (though I didn't cause I wanted it to last).

I've reseated the sb (had to use paint thinner to get the crap they use for tim off) and i'm running prime again. Unless this fixes it, I think i'll be buying this block, unless someone thinks I shouldn't. Anyone think I should buy the mosfet coolers too? They don't get very hot.

The alternative is a new motherboard. I'd prefer not buying one till I decide to CF, but if it's limiting me i would consider it.

 

[saint19]

lower cpu frequency? that would defeat the purpose... I just want to know if cooling the nb can actually be useful.

I don't think so, what error code do you get at the BSoD?

 

[slicedtoad]

here's the data from the event viewer:
- System
- Provider
[ Name] Microsoft-Windows-Kernel-Power
[ Guid] {331C3B3A-2005-44C2-AC5E-77220C37D6B4}
EventID 41
Version 2
Level 1
Task 63
Opcode 0
Keywords 0x8000000000000002
- TimeCreated
[ SystemTime] 2012-03-06T19:32:54.292404900Z
EventRecordID 39991
Correlation
- Execution
[ ProcessID] 4
[ ThreadID] 8
Channel System
Computer dmwin7
- Security
[ UserID] S-1-5-18
- Event Data
BugcheckCode 292
BugcheckParameter1 0x0
BugcheckParameter2 0xfffffa8007b0e028
BugcheckParameter3 0xbe200000
BugcheckParameter4 0x5110a
SleepInProgress false
PowerButtonTimestamp 0

I'll check what it shows on the actual bsod next time.

The thing that's weird is it's always around 8 hours after starting prime. I'm going to run memtest to rule out memory.

edit: the error code 0x8000000000000002 stands for "unkown bsod" right?

 

[rubix_1011]

Just sounds like an instability either from too much voltage, or not enough. Voltage+clock speed = more heat, so it's best to minimize voltage needed for a specific overclock. I'd say bump your voltage up a bit and retest...then move it down until it's unstable...this will help you find your happy place.

I doubt this has anything to do with temps causing the crashes...it's instability.

 

[slicedtoad]

Ok, i just ran a bunch of tests. first i changed some bios settings. can't restart right now so i can't list which ones, i'll add them later. Basically the options regarding amp limits, overvolting at load and vrm power limits have all been turn to enabled/extreme/wtvr.

here are my results:

bios vcore.....cpuz vcore at load........Result
1.400...........1.416v .........................crash after 2min
1.405...........1.424v .........................crash after 8min
1.410...........1.432v .........................crash after 35min
1.415...........1.440v .........................crash after 3 hours
1.420...........1.448v .........................crashed after 3 hours frozen rather than bsod
1.425...........1.456v .........................crashed after 3 hours
1.430...........1.464v .........................crashed 1 and 1/2 hours
1.435...........1.472v .........................Been running 10 hours so far.

I'll keep testing all night but hopefully I found a stable voltage and won't need any more parts. My max cpu core temp so far during the 10 hour run is 74C with an average of 63.1C across all cores and 66.9C on the hottest core.

 

[slicedtoad]

Best answer selected by slicedtoad.

 

[saint19]

This topic has been closed by Saint19